Manuscript Outline
Cram
Title:
Abstract
Models and observations suggest that that particle flux attenuation is lower across the mesopelagic zone of anoxic environments compared to oxic ones. This attenuation is likely a function of microbial metabolism, as well as agregation and disaggregation by zooplankton. Analysis of particle size spectra provide insight into the relative roles of aggregation, disaggregation and remineralization.
We measured particle size profiles at a station in the core of the Eastern Tropical North Pacific Oxygen Minimum Zone (ETNP OMZ) using an underwater vision profiler (UVP) multiple times of day, at different times of day, over the course of a week. We normalized our UVP measurements by comparing them to particle flux measurements measured by sediment traps. We also explored how our measurements related to acoustic observations of migratory marine species. We also compared our observations to UVP measurements from a site at a similar latitudes but with non limiting oxygen concentrations.
Particle numbers and size distributions showed a non monotocinc trend, with sharp decreases in particle numbers the photic zone and the lower 500m of the OMZ, but increases in the top 350 m of the OMZ and below the OMZ. These increases in particle number generally concurred with increases in the particle size distribution slope suggesting that the increase in particle numbers were due to a production of small particles.
Particle flux at our site was characterized by rapid attenuation in the top layer of the OMZ followed by either low attenuation or a small increase in abundance, depending on the day of the study, around 500m. This region coresponded with the presence of migratory plankton that spent the day in this region.
A model of particle remineralization and shrinking was used to diagnose whether particle size patterns violated the assumption that particles only sink and remineralize, and are neither aggregated, disaggregated, or tranported by zooplnkton. Our model identified aggregation like processes between 250 and 500m of in the water column that occurred at all time-points.
Our data suggest a role of zooplantkon in transporting biomass in the form of fecal pellets, into the core of the OMZ, but also in disaggregating particles in this same region. We further observe that there is temporal varibility in flux transport, but that this variability is small, only accounting for ~X % of the variability in flux, and not accounting for variability in particle numbers or size distribution over the one week time period studied.
Introduction
And:
The biological pump’s interaction with OMZs interaction is consequential
Pelegic oxygen minimum zones are expanding(Stramma et al. 2008), and this expansion is likely to effect ocean chemistry, the habitat of marine organisms, and the interactions between between organisms and chemistry (Gilly et al. 2013). Oxygen minimum zones have the potential to enhance carbon transport to the deep ocean, by inhibiting microbial degredation of sinking marine particles (Cram et al. 2018). However, biologial organic mater transport is modulated by zooplankton (Steinberg et al. 2008; Steinberg and Landry 2017) which feed on, produce and disaggregate particles, and whose interactions on particle flux in pelegic OMZs are only beginning to be explored .
Sinking particles and transport of organic matter by zooplankton are two main carbon gets transported from the surface into the deep ocean, and the processes that regulate this biological pump regulate global atmospharic carbon levels (Kwon).
This process is known as the biological pump (ref)
Several factors modulate this flux and include: Particle size, the dependence of particle degrading microorganisms on temperature and oxygen concentrations, the dependence of particle sinking speed on water viscocity and temperature, faster particle sinking due to ballasting by heavy materials such as carbonate and silicate, particle compositoin affecting breadown rates – for instance carbonate has been suggested to protect surrounding POM from degredation by bacteria alternatively particles in some region, or some kinds of particles may be more resistant to microbial degredation due to differences in POM content particle shape and porosity may also affect breakdown and sinking
- Zooplankton modulate organic mater flux this flux by repackaging particles (steinberg and refs therein), feeding at the surface and excreeting at depth (bianchi 2013 and others), thereby facilitating flux of carbon, and dying at depth (Steinberg and refs therein).
Zooplankton can further modulate flux by altering the particle size distribution. They can consume particles of a rane of sizes and repackage them into more uniform, perhaps faster sinking fecal pellets. Conversely, through sloppy feeding they can break large particles into multiple, slower sinking, smaller particles. {merge with point above}
Oxygen minimum zones are expanding (ref) and have been postulated to play a role in changing biogeochemical cycles.
They have unique effects on zooplankton because they limit the ability of anoxia sensitive zooplankton to migrate (ref) and also limit the ability of some anoxia sensitive predators. The limiting oxygen enviornment may also impact the behavior of zooplankton that are there as they avoid oxygen requiring behavior such as feeding. Community structure is distinct in the OMZ and characterized by XXX (Maas).
Our site
The ETNP is the biggist baddist OMZ (ref). While many OMZ, especially coastal ones, are characterized by high productivity which leads to carbon flux which removes oxygen from mid water depths (ref), the bulk of the ETNP OMZ is characterized by low productivity , with the low oxygen caused by slow ocean circulation through the OMZ region (refs clara sent). Many previous charicterizations of flux at this site, happened in coastal regions (that paper I used in my talks). Flux has not been previously characterized in this region (unless meg gets her paper out first). The site is characterized by nitrogen loss processes by both anammox and cannonical dentitrification processes. (Fuchsman cyano paper, probably many others).
Predicitons
Bianchi weber
- Models suggest that organic mater flux into the deep ocean is more efficient in oxygen minimum zones (Cram, Deveries and Weber). ** Bianchi and Weber proposed three reasons that this could occur
- Slower attenuation of all particles: The rate of remineralizatoin of all particles could be slower in OMZs.
- Decreased disaggregation by zooplankton:
- Slower attenuation of large particles: Large particles might harbor cores that are limited in both oxygen and nitrate and so microbial metabolism could be unually slow in these.
The authors proposed that these processes would have signature effects on particle size distribution in the core of the ETNP. Slower attenuation of all particles, was predicted to result in an increase in the abundance of small particles, while the other two models, would result in a decrease in small particle abundance as small particles were either not replaced by breakdown of large particles (model 2) or as those particles were broken down more quickly than larger particles (model 3). but they didn’t have data to determine which model was bestist.
Kiko
Zooplankton transport has also been postulated to be effected by zooplankton. Kiko et al. explored zooplankotn distributions and particle size distributions in the eastern tropical north atlantic, which has a weak oxygen minimum zone. They found transport into the omz core by zooplankton. They postuated, that this was possible becaues there was still oxygen. but similar analysis of particle size distributions and estemates of flux haven’t been done in the OMZ core. (kiko).
Methods Background
Underwater vision profilers have provided important insights into particle size distributions in the ocean. Flux profiles measured by traps can be compared to UVP measurments to estimate the size to mass and sinking speeed relationships of particles (ref). They can show useful things about spatial (guidi, kiko A) and temporal (kikoA and B) variability in the ocean. Indeed these data were critical in constraining a model (cram).
Traps tell us a lot about flux.
Acoustic data can tell us a lot about nekton distributions, and even those of plankton.
The meausurements have synergistic effects, as demonstrated by Kiko et al in the ETNA (damn them).
But
- Weber and bianchi’s model hasn’t been validated against real data.
Particle size has not been normalized to observed flux in this region.
The temporal dynamics at diel time scales have not been examined in the ETNP, or any other truly anoxic open ocean region. They have also never before been related to diel migration patterns.
It is not known how migratory zooplankton affect particle size distributions – either by tranport of particles, nor by disaggregaton processes.
It is not fully known how oxygen limitation affects particle size dynamics and flux.
Acoustic data, traps, and UVP measurments have not been combined in this region.
Therefore
We conducted the first ever time-series analysis of particles, in the oligotrophic region of the ETNP oxygen minimum zone. With the goal of normalizing to observed flux, and comparing particle size data to acoustic data about the distributions of zooplankton and larger orgasnisms.
We compare this data to acoustic signature of diel migrators, of mulitple size classes
We modified the a particle remineralization and sinking model (PRiSM) (Devries, Cram) to estemate particle size distribution changes that would occur under dynamics where particles only sink and remineralize, and compared observed distributions to ones expeced by the model. This allowed us to quantify the magnitude of non-PRiSM like processes.
We compared our results to those seen at a station at similar latitude in a non limiting region of the tropical north pacific.
Scientific questions:
- How do the particle size distribution at one location in the oligotrophic Eastern Tropical North Pacific evolve with respect to depth, and how does it vary over time?
- Do our data support any of the Bianchi-Weber models?
- Do our data suggest regions of the oxygen minimum zone with disaggregtion like processes, and if so do these co-occur with regions suggested to harbor zooplankton.
We hypothesized
- Temporal day to day variability in paticle number, particles size distribution slope and flux would be evident.
- This variability would relate to the location of migratory zooplankton, with a combenation of increased particle flux and disaggregation present where zooplankton occur.
- Disaggregation and particle production by zooplankton might lead to particle size patterns that cannot be explained by remineralization and sinking alone.
- All three of the Weber Bianchi hypotheses.
Methods
Most measurments were taken on board the R/V Sekuliaq from 07 January 2017 thorugh 13 January 2017 at XXX Lat XXX Long. Data are compared against {describe station 100 here}
Water property measurment
We measured water properties of temperature, salinity, fluorescence, oxygen concentration and turbidity using the shipboard XXX CTD {get sensor information}. Data were processed using seabird software and analyzed and visualized in R.
Particle size measurments
Particle size data were collected by underwater vision profiler (UVP) that was mounted below the CTD-rosette and deployed for all CTD casts shallower than 2500m. A UVP is a combenation camera and light source that describes the abundance and size of particles from 100 microns to several centemeters in size (ref). Particles have been previously shown to be primarely “marine snow” but may also include a small number of zooplankton and visual artifacts. UVP data were processed using custom matlab scripts, uploaded to XXX, and analyzed in R.
Flux measurments
Particles were collected in incubating particle traps (REF). Traps were used to performincubation studies which will be reported elsewhere. As part of these studies, the traps also generated data about carbon flux, which is reportd here. Two types of traps were deployed. The particles were collected in two kinds of traps. One set of traps, generally deployed in shallower water had a solid cone opening with a X m diameter cone opening. The second set had larger conical Xm diameter 200 micron nylon mesh net at the top. In all cases particles collected in the net or cone fell into one of two chambers. The “plus-particles” chamber collected particles from the net and incubated them for an amount of time that ranged from X days to X days. The top-collector trap collected particles, and then returned immediately to the surface. We prferentially used data from the “top-collector”; however in many cases, data was only available forom the “plus-particles” trap, in which case we used that data.
{What if I move the analysis to the results section} ## Analysis
The total number of particles in each size were visualized. We calcualted total particle number by multiplying over all size bins.
Particle size distribution
We determined the slope and intercept of the particle size distribution spectrum by fitting a power law of form \(log(# Particles) = Int + PSD * log(Size)\). Because large particles were infrequently detected we used a poisson- general linear model that considered the volume of paricles sampled and that assumed that the residuals of the data followed a poisson (rather than normal) distribution.
Estimating particle flux
We estemated particle flux by assuming that particle flux in each size bin followed the equation flux = C_f * d ^ a (guidi 2008 ref). We used Alledredge and XXX’s estemate of a and then fit the observed flux measurments to particle measurments to estemate C_f. We then used this equation to estimate flux from the data collected by the UVP.
Size specific information
We seperately analyzed total particle numbers, particle size distribution, and particle flux for particles larger than 53 microns, and those smaller than 53 microns to determine the relative contributions of these two particle classes to particle properties
Variability
We quantified whether the sample to sample variability
Model Formuation
Results
Physical and Chemical Data
The anoxic zone, characterized by undetectable oxygen levels extends from 80 m to 850 m depth, with a sharp upper oxycline and a gradual lower oxycline (Figure 1B-D). The upper oxycline tracks a sharp picnocline (Figure 1C 1D), characterized by a abrupt drop in temperature below the mixed layer, and an increase in salinity (Figure 1B). The site is characterized two fluorescence maxima (Figure 1C). The larger, shallower fluorescence peak is positioned just above the oxycline, ending exactly where oxygen reaches zero. The smaller, lower peak is positioned entirely inside of the anoxic zone. Turbidity tracks the two chlorophill peaks in the surface, and has a tertiary maximum at the lower oxycline (Figure 1D).
Acoustic data reveal diel migration patterns
Acoustic data, produced by the shipboard EK60 , suggest the presence of multiple cohorts of migratory organisms. The largest organisms, observed by backscattering of the EK60’s 18000 Hz signal, showed the clearist patterns. Most migratory organisms apppeared to leave the surface at dawn and return at dusk, spending the day between 250 and 500m (Figure 2A). There appeared to be two local maxima in backscattering intensity at mid-day, one at ~300m and one at ~375 m (Figure 2A). There also appeared to be organisms that migrated downward at dusk and upward at dawn , spending the night at ~300m (Figure 2B). There was also a peak of organisms that appeared, at mid-day, on some but not all days, without any visible dawn or dusk migration just above the base of the photic zone. (Figure 2C). Other characteristics included what appeared to be diel migrators that crossed the OMZ and spent the day below the range of the EK60 (Figure 2D), as well as organisms that appeared between 500m and 1000m but did not appear to migrate to or from that depth at our site, but rather simply transeted through the the EK60’s field of view at that time (Figure 2D).
Similar patterns were evident each other measured frequency, with better resoultion by the lower frequencies (Figure S_).
Flux data
Flux measurmenets at station P2 were consistant between the different particle trap types and chambers measured, and showed a profile that broadly represented a power law with respect to depth, with the exception that flux appeared to increase in one trap at 500m. Four traps in the surface had anomalously low measurments of flux, compared to similar traps placed at similar depths, which may have been due to trap malfunctions {Talk to Jaqui about these}.
Particle abundance measurments vary with size and depth
In all profiles, particle abundances were highest at the surface, and highest among the smallest particles (Figure S_). Visual examination of the relationship between particle number and size suggested a power law relationship where the log of volume and binsized normalized particle abundance was proportional to the log of the particles size (Figure S_). The exception to this pattern was very large particles, which are rare enough that they are usually not detected by the UVP. Generalized linear models that assume a negative binomial distribution of the data were able to account for this undersampling of large particles estemate power log slopes, while taking into account rare occurances of the data, at each depth (Figure S_).
Total particle numbers were generally similar between different casts, regardless of which day or hour they were collected (Figure S_). Particle numbers were highest in the surface and decreased rapidly, flattened out over the 250 m to 500m range, decreased again untill the lower oxycline, and then increased below the oxycline (Figure S_).
The particle size distribution slope steepened (became more negative) between the surface and 500m, flattened (became less negative) between 500m and 1000m, and then steepened again after 1000m (Figure X, S_). Steeper, more negative, slopes indicate a higher proportion of small particles relative to large particles, while flatter, less negative, slopes indicate a higher proportion of large particles relative to other places.
Estemated particle flux sometimes increases with depth in the OMZ core
Using an optimization algorithm, we found that there was greatist agreement between estimates of trap observed particle flux, and UVP estemated particle flux when the particle size to flux relationship was goverened by the ratio \[Flux = 133 * Size ^ {2.00}\]. This resulted in a UVP predicted flux profile that broadly fit the expected trap observed flux profiles, excluding the four traps that were held out from the analysis due to low abundance.
Particle flux profiles varied notably between casts between the base the photic zone and 100 and 500m m (Figure 4a-b). Between 250 m and 500 m particle flux appeared to increase on some but not all casts, while attenuating slowly on others (Figure 4c). Below 500m, there were not enough casts to measure variability between casts.
General aditive models that examined the rate of change of flux between 250 and 500m found that, after removing the effect of depth, there was a statisticlaly significant relationship between day of the week and the fifth-root transformed, rate of change of flux (P = 0.002), as well as between hour of the day and flux (P = 0.040). There were increases in flux over this region towards the beginning and end of the sampling period, and lowist near day 10. There was also increases in flux in the daytime but decreases at night-time. By comparing three general additive models, one that considered only depth, one that considered depth and day of the week and on that considered depth, day of week, and time of day, we found that while depth accounted for 37% of the variance, adding day of the week accounted for an additional 18% of the variance, and hour of the day accounted for only 8.7% of the remaining variance in transformed rate of change of flux. If the fifth root transformation was not applied to the rate of change of flux, the hourly pattern was not evident. Increases in flux in this region were clearly not limited to the daytime, as one midnight cast showed increases here as well (Figure 4C).
ETNP particle dynamics differ from those seen at an oxic site
{I need to add these supplemental figures}. ### Oxic Region: Chemistry The oxic site was characterized by a more gradually picnocline, and an oxygen minimum at 500m that was not fully anoxic. The photic zone characterized by a single fluorecence peak with a maximum at 110m and which disapeared at 200m. Turbidity followed chlorophyl and did not have a deep peak. There was a salinity peak at 150m.
Oxic Region: Particle Number and Size
With the exception of the surface, where particle numbers were similar between the two sites, particle numbers were higher throughout the top 1000m of the water column at the ETNP site, than at the same-latitude, oxygenic P16 station 100. Particle size distributions were similar between the two sites above 500m, being characterized by overlapping confedence intervals by generated by a general additive model. From 500 to 1000m, particle size distributions were steeper at the ETNP site,being characterized by a higher proportion of small particles (Figure S5).
Oxic Region: Large vs small particles
Small particles (100um - 500 um) at our site were about two orders of magnitude more common than large particles (>= 500um). The number of large particle numbers appeared to attenuate more quickly than small particles, and more genearally follow a power law decrease, while small particles appeared to increase around 500m (Figure S_). Flux was predicted to be predominantly from small, rather than large particles, at all depths except the very surface. The particle size distribution, calculated only on large particles, was more variable between depths than calculated for small particles. Data from the oxic P16 station 100 suggested more particles, steeper particle size distribution, and more flux than at this station than at the ETNP station. They also sugested that differences between large and small particles, with respect to number, flux and size distribution that were broadly similar to the ones seen at ETNP staion P2 (Figure S_).
Flux
Flux at P16 Station 100 appeared to attenuate following a power law from the base of the photic zone through 1000m (Figure S_ A&B).
Smoothed and averaged data
Highly smoothed particle data suggestd that particle size, averaged accross all casts, followed a pattern in which the abundance of small particles increased in the OMZ surface (Figure 5A), which corresponded with characterized by steepening of the particle size distribution (5A), an incrrease in small particle biomass (5B), but not of large particle biomass (5C). Deeper in the OMZ the small particle number, PSD slope, and biomass of small particles declined.
Particle number dynamics differ from model expectations
We were able to use our modified particle remineralization and sinking model to predict particle size distributions at each depth from the particle size distribution at depth one depth-bin shallower and the calculated flux attenuation between the two depths. We found that the observed particle size distributions usually varied from model expectations (Figure S_ {An example}). Tautologiclly, at each depth, the observed size profile and the model predicted size profiles have same flux. However, the difference between the flux of observed and predicted small particles (100-500), normalized to depth, serves as a valuable metric of patterns of deviations from modeled results. We call this value OSMS (Observed Small Flux Finus Modeled Small Flux).
\[ OSMS = \frac{(Small\,Flux\,Observed - Small\,Flux\,Modeled)}{\Delta Z}\] In the above equation \(\Delta Z\) is the distance, in meters between the current depth, and the depth for which the particle size distribution and flux change since, inform the model.
OSMS was positive between the photic zone and 500m, meaning that less small flux attenuated than would be expected from the PRiSM model in this region. There was some variability in the OSMS parameter between casts. A general additive model, after factoring out the effect of depth, found that there was a statistically significant relationship between day of the cast and OSMS with highest values near day 10 of the study (which is when flux attenuation in this region was lowest) (p=0.01). However there was not a statistically significant relationship between hour of the day and OSMS.
Below 500m, OSMS was negative. There were only two casts that reached below 500m at this station, and so an analysis of the dynamics of OSMS in this region are not possible.
At P16 Station 100, OSMS was positive between the base of the photic zone and 350m and negative below 350m (Figure S_).
Discussion
Diel migrators spend time in the OMZ
Organisms of all sizes appear to migrate into the core of the OMZ. Most in the day Figure 2A, and some at night Figure 2B. The leading hypothesis is that the OMZ provides protection from predators of all sizes. Other studies with mocness have characterized the diurnal and nocturnal orgnaisms found in the OMZ, though not at this exact site. In the XXX, maas et al saw decapods and stuff {reread that paper and find others}. At the size range seen by the 18000 kHz band, humboldt and market squid are known to be tolarant to anoxia and to migrate into the OMZ to hunt. These small organisms likely have the greatist effect on particle transport (ref) and disaggregation (ref). Diel migration patterns are common in the ocean (ref) and zooplankton transport of particles have been indicated in other enviornments (ref).
The acoustic data suggests the presence of some other interesting migratory patterns. The organisms that appear between 500m and 1000m may be jellyfish (ref). The horizontal bands seen here could indicate the presence of swarms of jellyfish that migrate through the field of view of the EK60. These organsisms do not seem to have much of an effect on particle size.
Flux is lower at this site than previous measurments in the ETNP {Unless Megan Duffy is going to have already said this somewhere}.
Flux here is lower at all depths than seen in previous measurments by traps in the OMZ (Ref). This seems reasonable, because the previous measurments were taken nearer to the coast, where surface chlorphyl is higher (Figure 1A).
The flux to size relationship is typical of other sites.
The exponent particle size to flux relationship that we saw at our site 2.00, is of a similar magnitude to those seen by other studies that compare UVP flux to traps (Guidi, Kiko). It is not identical to these measurments. This could be because these values vary between sites, or that imprecision in flux measurments leads to differences in these values between studies. Indeed, we found this value was sensitive to outlying data points.
If we left in the four traps that measured very little flux in the surface, we instead got values for this size relationship that approached zero. We feel confedent excluding these traps, because it was standard for traps to under-measure flux {and maybe there were actually things observed about these}, and because we have traps at very similar depths over the same time window that provided substantially higher results. Because we have found that traps appear to under-measure flux when they fail, rather than over measure it, we have gone with the higher measurements.
Remineralization rates of all particles decrease in the OMZ, but disaggregation does not
Particle size profiles, particle size distribution slopes, and estemated biovolume, averaged accross all casts and smoothed, are all similar to the predictions made by Weber and Bianchi’s (YYY) “Model 1”. (Figure 4). This suggests that the low oxygen at this site decreases the particle remineralization rate of all particles, including small ones. It does not support the Weber-Bianchi Model 2 in which remineralization is suppressed in the OMZ, nor their model 3 in which only the very large particles’ remineralization is slowed.
Zooplankton likely transport organic mater into the OMZ core.
Predicted flux levels sometimes increase between XXX and XXX meters, and at all time attenuate very slowly in this region. The EK60 data suggest diel migration of organimss of all sizes to ths same region. Taken together, this increase in flux concurent diel migration, suggest transport of organic mater by zooplankton. Zooplankton may consume organic matter in the day and then release it at night (REF). That the flux varies between days suggests some day to day variability in this transport. That it is highest in the afternoon suggests a release of fecal pellets just as the plankton are returning to the surface. Perhaps they use fecal matter as ballast material, using it to aid in sinking and remaining in the deep ocean and then releasing it just as they need to migrate to the surface, where they become positively bouyant.
Zooplankton likely disaggregate particles in the OMZ core.
The observation that there is more small flux that would be predicted by remineralization and sinking between the photic zone and 500m suggests that some process is disaggregating large particles into smaller ones. That this corresponds with the region where migratory organsims are found suggests that some of these organsisms, likely zooplankton including copepods (ref), may be breaking down particles through “sloppy feeding” in which organisms attack particles and break them into smaller pieces (ref). Alternatively, other processes such as rapid spontanious or microbial breakdown of zooplankton transported particles, or of flux from the surface, could also be responsible of this increase in small particle flux (ref).
Other deviations from model assumptions could also explain the increase in small particles over model predictions For instance, if the models assumed relationship between size, flux, sinking speed and biomass are not all accurate, particle dynamics would also differ. For instance, if remineralization differed between particle types, with small particles breking down more slowly than larger ones, perhaps because they are more labile, we might see the same kind of deviation from the model. If small particles sank more quickly for their size than expected, as has been seen elsewhere {Mcdonnell}, a similar deviation would occur as they would have less time to remineralize per depth.
Our model also assumes a spherical paricle drag profile, such that the particle sinking speed fractal dimension (gamma) is one less than the particcal size fractal dimension (alpha), and that these two values sum to the particle flux fractal dimension. If any of these assumptions do not hold, or even if our calculation of the particle flux fractal dimension was in error, the magnitude of the values may differ.
Furthermore, since flux varies over time, that variability could contribute to these values, and we haven’t yet deconvoluted these processes, but future models could leverage time series data like this one to incorperate multiple observed time-points into the prediction of particle size distributions at depth. One way to do this would be to use a smoothing function to interpolte particle abundances at each size, depth and times, and then to use a model in which the sinking speed of a given particle size is used to identify the relevant time-point where the abudance informs that time-point.
The opposite pattern to that seen at shallower depths occures below 500m, with apparent flattening of the particle size distribution. This could suggest aggregation occurs here, though given the sparsity of particles, we don’t see a mechanism for this process. More likely is that particles do follow PRiSM like processes in this region but that likely one of our parameters are off and so disaggregation is actually higher than shown in this figure.
Figures
Figure 1. Overview of the geography, physics and chemistry of ETNP station P2 (A) Map of the ETNP Oxygen Minimum Zone and the location of station P2. Colors indicate chlorophyll concentrations at the surface, while the red ouline signifies the region containing low oxyegen. The red circle indicates the location of Station P2. (B-D) Oceanographic parameters collected from a cast at 2017-01-13 12:15 CST (local time). All profiles contain a plot of oxygen concentrations. When available, the thin horizontal green line shows the location of the base of the photic zone (160m), while the horizontal blue line shows the base of the oxycline. Figures B and D also show density (Dashed Gray Line). (B) highlights temperature and salinity. (C) fluorescence, focusing on the top 300m of the water column, and (D) beam attenuation.
Figure 2. Acoustic data, measured by EK60, measured over the course of the experiment. Shown are data from the 18000 Hz frequency band, which have highest depth penetration, but which appear to co-occur with data from other frequency bands (see Figure SX). Values are in return signal intensity and have not been normalized to observed biomass. Several interesting pattens can be seen. A. Two bands of organisms can be seen leaving the surface at dawn, spending the day between 250 and 500m and returning to the surface at dusk. B. Another group of nocturnally migrating organisms can be seen leaving the surface at dusk, spending the night near 250m and returning at dawn. C. Some organisms appear at the base of the photic zone, during some, but not all mid days, and then disapear in the evening. D. A group of very deep migrating organisms appears to leave the surface with the diel migrators and pass all the way through the OMZ and out of the EK60’s field of view. It returns at dusk. E. Swarms of organisms apear between 500 and 1000m disapearing later in the day. Swarms apear in the deepist layers at night and appear progressively shallower as the day progresses. {ADD LETTER LABELS}.
Figure 3. Particle flux, measured from sinking traps large symbols. Data from the “plus particles” and “top collector” samples from both cone and net traps were collated to generate these data. Trap types are shown by the shape and color of the large points. Superimposed are binned estemates of particle flux generated by fitting the sum of particle numbers all four profiles, binned as in Figure X, to the trap observed flux. The four points enclosed by the rectangle are unusually low compared to other traps collected at the same depth, and were therefore excluded from the fit. {Convert UVP points to a line, maybe leave points where the traps ares}
Figure 4. Within and between day variability in UVP predicted particle flux at ETNP station P2. Profiles are compared against P16 station 100, a non OMZ station at similar latitude in the tropical pacific. All profiles are depth binned with higher resolution towards the surface (methods). (A) Flux profiles in the top 1000m of the water column. (B) A more detailed depiction of the area enclosed by the rectangle in A. (C) The rate of change of flux, devided by the rate in change in depth. We show the fifth root of these values in order to highlight differences between values close to zero.
Figure 5. (A) Gam smoothed binsize and volume particle numbers at each particle size class. (B) Particle size distributions. And estimated biomass of (C) Small and (D) Large particles. {I need to get rid of the green and blue line at P16 – and maybe calculate the photic zone therein}.
Figure 9. Quantification of non remineralization and sinking like processes. Points indicate the difference between the observed small particle flux, and the flux that would be estimated if particles from the size distribution in the depth bin above remineralized and sank only following the PRiSM model. Values are normalized to the change in depth. Thus values are uMol Carbon/m3/day {change to “Deviation from Model”, keep units}
Supplemental Figs
[move to supplement]
Figure S1. Acoustic data, measured by EK60, measured over the course of the experiment. Shown are data from the all frequency bands. Values are in return signal intensity and have not been normalized to observed biomass.
Figure S2. A profile of particle abundances at different sizes and depths. (A) Numbers of observed particles and (B) particle numbers normalized to volume sampled and particle size bin width. (C) Smoothed and extrapolated particle abundances, based on a negative bionomial gam that predicts particle abundance form size and depth.
Figure S2. An example of observed particle size distribution spectra. These are depth binned data from between X and X m deep in the water column from the cast that occurred at DATETIME for stn_043. A total volume of XXX L of water are sampled herein. Points indicate (A) total numbers of observed particles and (B) particle numbers normalized to volume sampled and particle size bin width. The line indicates the predicted best fit line of the data. The line was fit on the bin and volume normalized data by a negative-binomial general linear model. The line in panel A indicates predictions from this same model, rescaled into absolute particle space.
Figure S3. (A) Observed, volume normalized total particle numbers from 9 casts taken at different times of the day at ETNP station P2. (B) Calculated particle size distribution slopes of those particles. These data have not been binnedby depth.
Figure S4. Physical and chemical data from P16 Station 100. Located at 16.5°N 152.0°W. (A) Map of the nearby tropical pacific station P6 Station 100. Colors indicate chlorophyll concentrations at the surface, averaged over all MODIS images. The red circle indicates the location of Station P2. (B-D) Oceanographic parameters. The thin horizontal green line shows the location of the base of the photic zone (200m m). A Oxygen, and fluorescence. Because the fluorometer was broken on this cruise, fluorescence data were pulled from world ocean atlas. B Oxygen temperature and salinity. (C) Beam attenuation and density, calculated from the salininity temperature and pressure data.
Figure S5. As above, but for the final cast taken at ETNP station P2 and the only cast collected from the P16 transect at the station 100. P16 Station 100 was chosen because it is at a similar latitude to ETNP station P2. (A) Total particle numbers, (B) Particle size distribution. {Cut to 1000m}
Figure S6. Flux profiles and flux attenuation at P2 Station 100. (A) Flux profile (B) Fifth-root transformed depth normalized rate of flux decrease. (C) Difference between observed and modeled results. Higher values suggest more disaggregation-like processes.
Figure S_. Gam predicted effects of A Depth, B Day of the month in January 2017, and C hour of the day on the fifth-root transformed, depth normalized, rate of change of flux. Y axis indicates the value of the component smooth functions effect on Flux. Positive values associate with times and regions of the water column where flux is increasing, holding other factors constant, and negative ones where it is decreasing.
Figure S_. Gam predicted effects of A Depth, B Day of the month in January 2017 Y axis indicates the value of the component smooth functions effect on the difference between observed and modeled flux. Thus higher values correspond with greater flux of small particles than predicted by the model.
Figure S6. Depth binned particle number (volume normalized), particle size slope (psd), and flux (estimated as in Fig. 4) for large (\(>= 500\, \mu m\)), small (\(< 500 \, \mu m\)) and total particles, at the oxic and anoxic site
#MOVED Maybe skip? {done, notadded} Figure S1. A profile of data generated by the UVP. At each depth the abundance of particles at each size are color coded on a log scale. Particle sizes where no particles in that sample were seen are represented with a smaller black dot.
{done not added} Figure S2. Comparason of total particle number and particle size distributions of all casts taken at the ETNP station P2. Points indicate individual samples, while ribbons indicate confedence intervals of those sampeles. The overlapping confedence intervals suggest that there is not a statistically detectable difference between the different casts at the same station. {I’d like a more quantitative metric}.
References
Cram, Jacob A., Thomas Weber, Shirley W. Leung, Andrew M. P. McDonnell, Jun-Hong Liang, and Curtis Deutsch. 2018. “The Role of Particle Size, Ballast, Temperature, and Oxygen in the Sinking Flux to the Deep Sea.” Global Biogeochemical Cycles 32 (5): 858–76. https://doi.org/10.1029/2017GB005710.
Gilly, William F., J. Michael Beman, Steven Y. Litvin, and Bruce H. Robison. 2013. “Oceanographic and Biological Effects of Shoaling of the Oxygen Minimum Zone.” Annual Review of Marine Science 5 (1): 393–420. https://doi.org/10.1146/annurev-marine-120710-100849.
Steinberg, Deborah K., and Michael R. Landry. 2017. “Zooplankton and the Ocean Carbon Cycle.” Annual Review of Marine Science 9: 413–44. https://doi.org/10.1146/annurev-marine-010814-015924.
Steinberg, Deborah K., Benjamin A. S. Van Mooy, Ken O. Buesseler, Philip W. Boyd, Toru Kobari, and David M. Karl. 2008. “Bacterial Vs. Zooplankton Control of Sinking Particle Flux in the Ocean’s Twilight Zone.” Limnology and Oceanography 53 (4): 1327–38. https://doi.org/10.2307/40058255.
Stramma, Lothar, Gregory C. Johnson, Janet Sprintall, and Volker Mohrholz. 2008. “Expanding Oxygen-Minimum Zones in the Tropical Oceans.” Science 320 (5876): 655–58. https://doi.org/10.1126/science.1153847.